GREAT METEORITE COLLECTIONS. 159 



N. B. — It is, of course, to be remembered that after a collection 

 of meteorites passes the number of three hundred or three hundred 

 and fifty kinds, every added kind will — in almost every case — 

 materially reduce the average weight of the pieces forming the 

 whole collection. Here an excellence in one factor will be the 

 deterioration of another. 



THE PROPORTION OF SPECIMENS OF INTEREST IN THE GROWTH OF 

 THE SCIENCE ALSO TYPE SPECIMENS. 



In a collection of material representing any subject, there will 

 be some objects which mark points where new views of the subject 

 had their birth. With meteorites this has been a specimen showing 

 a crust, another showing pittings, another orientation, another 

 chondri, another Widmanstiitten figures, another new mineral 

 combinations, another alteration of structure by the addition or the 

 substitution of an element of its composition or by the apparent 

 conditions of its origin. 



With the observing of each of these features has come an 

 added growth in the science itself. Hence the value of these 

 growth-registering specimens. The number of these is not large, 

 and they are distributed in many collections. There they exist as 

 type-specimens of high interest and value. 



THE PROPORTION OF SPECIMENS SHOWING LEADING POINTS IN METE- 

 ORITE CLASSIFICATION, EITHER CHEMICAL OR PETROGRAPHICAL, 



The former factor noted features which were discovered seriatim 

 and progressively. The present one results more from close study 

 at a later period of large series of specimens. In this many 

 divisions are formed, each one based upon a specimen of a definite 

 composition or an especial structure, which thus becomes a type 

 of its kind. Again, a specimen newly studied comes to modify or 

 to destroy a plan of classification previously adopted. In all this, 

 as in the preceding division, the original specimens have the very 

 highest value as type specimens, to be forever preserved and 

 referred to as controlling types. 



Two great collections — Vienna and Paris — possess these in the 

 main, and thus far they are unapproachable by others. Vienna, 

 from Partsch's first essay of classification in 1862 to the present 

 day, has built taxonomical structures ; the last of these — that of 

 Brezina, 1904 — being composed of 74 groups. Paris classifications 

 were led by Daubree in 1867. Meunier has been more prolific — with 



